MAST CELL RESPONSES TO DANGER SIGNALS

Transcription

1 Department of Medicine, Clinical Immunology and Allergy Unit, Karolinska Institutet, Stockholm, Sweden MAST CELL RESPONSES TO DANGER SIGNALS Mattias Enoksson Stockholm 2012

2 All previously published papers were reproduced with permission from the publisher. Paper I: Copyright S. Karger AG, Basel Paper II: Copyright The American Association of Immunologists, Inc. Published by Karolinska Institutet. Printed by Larserics Digital Print AB Mattias Enoksson, 2012 ISBN

3 A problem solved is a problem caused - Karl Pilkington

4

5 ABSTRACT Detecting and responding to danger is a paramount function of the immune system. Compounds heralding danger can be divided into two groups: exogenous and endogenous danger signals. The former group consists of conserved microbial structures such as lipopolysaccharide (LPS), while the latter consists of host compounds released or exposed by dead or dying cells as a consequence of trauma, stress or infection. Mast cells are long-lived immune cells present in almost all tissues, and are especially numerous at sites facing the external environment, making them ideal responders to danger signals. The aim of the work presented in this thesis was to investigate mast cell responses to danger signals of exogenous and endogenous origin. In Paper I, we investigated mast cell responses to the exogenous danger signal M-TriDAP, a bacterial peptidoglycan degradation product. We found that cord bloodderived mast cells (CBMCs) express NOD1, the receptor for M-TriDAP. Furthermore, M-TriDAP-treatment of CBMCs resulted in degranulation-independent release of cytokines and chemokines such as TNF, IL-8/CXCL8, MIP-1α/CCL3 and MIP-1β/CCL4. Importantly, we observed an augmented response when M-TriDAP was combined with the TLR4 agonist LPS, indicating cooperation between intracellular and extracellular pattern recognition receptors. In Paper II, we investigated mast cell responses to cell injury by subjecting murine mast cells to the supernatant of fibroblasts rendered necrotic by freeze-thawing. We found that mast cells respond to cell injury in this model by initiating a proinflammatory response, characterized by degranulation-independent release of cytokines and leukotrienes. By using genetically modified mice and molecular inhibitors, we found that the recognition of cell injury was MyD88-, T1/ST2- and p38- dependent. Finally, by using RNA-interference, we could pinpoint IL-33 as the necrotic cell compound that was responsible for the mast cell activation. In Paper III, we investigated responses to IL-33 administration in vivo. Here we found that wild-type C57BL/6 mice respond to intraperitoneal IL-33 administration with neutrophil infiltration. This response was not observed in mast cell-deficient mice but could be restored upon mast cell reconstitution, thus demonstrating a mast cell dependent mechanism. In Paper IV, we investigated the hypothesis that mast cells might function as sensors of damaged epithelia by responding to IL-33 during chronic inflammations of the airways, for instance in asthma. We found that IL-33 is released from necrotic airway epithelial cells and that CBMCs respond to the necrotic supernatant of these cells by secreting IL-5, IL-8/CXCL8, TNF and GM-CSF. However, no release of histamine, LTB 4 or PGD 2 could be detected. Interestingly, the exact same mediator release pattern was observed when CBMCs were treated with recombinant IL-33, suggesting that IL-33 might be an important factor released by injured airway epithelial cells that activates mast cells. In conclusion, the work presented in this thesis provides further evidence for important roles of mast cells in innate immune responses. The function of mast cells as sensors of cell injury is highlighted; a role that potentially can be either beneficial or detrimental. Finally, novel evidence is provided for the notion that IL-33 is an important danger signal capable of mast cell activation.

6 LIST OF PUBLICATIONS I. Enoksson M, Ejendal KFK, McAlpine S, Nilsson G, Lunderius-Andersson C Human cord blood-derived mast cells are activated by the Nod1 agonist M-TriDAP to release pro-inflammatory cytokines and chemokines J Innate Immun. 2011;3: II. Enoksson M, Lyberg K, Möller-Westerberg C, Fallon PG, Nilsson G, Lunderius-Andersson C Mast cells as sensors of cell injury through IL-33 recognition J Immunol Feb 15;186(4): III. Enoksson M, Fallon PG, Lunderius-Andersson C, Nilsson G Intraperitoneal influx of neutrophils in response to IL-33 is mast cell dependent Submitted IV. Enoksson M, Andersson C, Heinz G, Chatterjee M, Saluja R, Erjefält J, Nilsson G Human mast cell responses to injured airway epithelial cells implications for chronic airway inflammation Manuscript

7 Publications not included in this thesis Lunderius-Andersson C, Enoksson M, Nilsson G Mast cells respond to cell injury through the recognition of IL-33 Front Immunol. 2012, In press. Review McAlpine SM, Enoksson M, Lunderius-Andersson C, Nilsson G The effect of bacterial, viral and fungal infection on mast cell reactivity in the allergic setting J Innate Immun. 2011;3: Review

8

9 CONTENTS 1 INTRODUCTION Mast cells Mast cell distribution and heterogeneity Mast cell activation and mediator release Mast cell function in health and disease Pattern recognition Toll-like receptors Nod-like receptors C-type lectin receptors Retinoic acid-inducible gene (RIG)-I-like receptors Damage recognition Endogenous danger signals IL IL-33 expression and signaling IL-33 activity IL-33 in health and disease IL-33 as a danger signal THE PRESENT STUDY Aim Methodology Results and discussion Human cord blood-derived mast cells are activated by the Nod1 agonist M-TriDAP to release pro-inflammatory cytokines and chemokines (Paper I) Mast cells as sensors of cell injury through IL-33 recognition (Paper II) Intraperitoneal influx of neutrophils in response to IL-33 is mast cell dependent (Paper III) Human mast cell responses to injured airway epithelial cells implications for chronic airway inflammation (Paper IV) FUTURE PERSPECTIVES POPULÄRVETENSKAPLIG SAMMANFATTNING ACKNOWLEDGEMENTS REFERENCES... 40

10 LIST OF ABBREVIATIONS AHR Airway hyper-responsiveness AIM Absent in melanoma APC Antigen-presenting cell ATP Adenosine triphosphate BMMC Bone marrow-derived mast cell CARD Caspase activation and recruitment domain CBMC Cord blood-derived mast cell CIA Collagen-induced arthritis CLR C-type lectin receptor DAMP Danger-associated molecular pattern DC Dendritic cell ELISA Enzyme linked immunosorbent assay GM-CSF Granulocyte macrophage colony-stimulating factor HMGB1 High-mobility group box 1 protein HSP Heat shock protein IFN Interferon Ig Immunoglobulin IL Interleukin IL-1RAcP IL-1R accessory protein IPAF ICE-protease-activating factor IRF Interferon regulatory factor LGP2 Laboratory of genetics and physiology 2 LPS Lipopolysaccharide LRR Leucine-rich repeat LT Leukotriene MCP Monocyte chemoattractant protein MDA5 Melanoma differentiation associated factor 5 MDP Muramyl dipeptide MEF Mouse embryonal fibroblast MIP Macrophage inflammatory protein NALP NACHT, LRR and PYD-containing protein NBD Nucleotide-binding domain NK Natural killer NLR NOD-like receptor NLRP NACHT, LRR and PYD-containing protein NOD Nucleotide-binding oligomerization domain PAMP Pathogen-associated molecular pattern PBS Phosphate-buffered saline PG Prostaglandin PGN Peptidoglycan PNEC Primary nasal epithelial cell PRR Pattern recognition receptor PYD Pyrin domain qpcr Quantitative real time PCR

11 RA RAGE RIG-I RLR SCF sirna TGF Th TIR TLR TNF Rheumatoid arthritis Receptor for advanced glycation endproducts Retinoic acid-inducible gene I Retinoic acid-inducible gene (RIG)-I-like receptor Stem-cell factor Small interfering RNA Transforming growth factor T helper Toll/IL-1R homology Toll-like receptor Tumor necrosis factor

12

13 1 INTRODUCTION The immune system has evolved to combat infections, a function that represents a prerequisite for life. Therefore, strict control and regulation of the immune system is paramount, as both excessive and insufficient activation can be fatal, leading to autoimmunity and persistent infections, respectively. Consequently, contextappropriate activation is vital [1], i.e. the nature, magnitude and duration of the immune response need to be appropriate for a given situation and location. In essence, the immune system can be activated by two means: i) by recognizing non-self molecules of infectious agents such as bacteria, virus, parasites and fungi (pathogen-associated molecular patterns; PAMPs) [2] or ii) by recognizing self molecules released by damaged or dying cells (danger-associated molecular patterns; DAMPs) [3]. To face the ever present threat of detrimental pathogens, the immune system has evolved into a highly organized network of organs, cells and molecules that together can be regarded as the barriers, first line-defenders and the heavy artillery of the immune system. Traditionally, the immune system is described to be composed of two branches: the innate and the adaptive branch. While the former is unspecific but fast in onset, the latter is slow but extremely specific. The first obstacles faced by invading pathogens are physical and chemical barriers, which can be considered part of the innate immune system. Such barriers include for example the skin, mucosal surfaces and low gastric ph. Should pathogens circumvent these barriers, the cellular part of the innate immune system will be activated. The cellular part is composed of a variety of cell types including macrophages, monocytes, dendritic cells (DCs), eosinophils, basophils, natural killer cells (NK cells), neutrophils and mast cells. Together, these cells possess the abilities to directly kill pathogens, produce compounds that enhance pathogen elimination and to activate the adaptive branch of the immune system. Most pathogen infestations can be handled by the innate immune cells, but sometimes a more forceful response is needed, and this is where the adaptive branch of the immune system is introduced. Antigen-presenting cells (APCs) of the innate immune system, such as DCs, are specialized in carrying antigens through the lymphatic system to local lymph nodes, where activation of the adaptive immune system takes place. The adaptive immune system is composed of T and B lymphocytes, cells that carry receptors specific for a vast amount of antigens. When activated, these cells expand and participate in pathogen killing by producing mediators and antibodies that enhance pathogen elimination. The adaptive immune system also possesses memory, which facilitates a fast response to reinvading pathogens. In order to fully understand the immune system, and eventually translate this understanding into therapeutical use, detailed knowledge of the various components of the immune system is of great importance. This thesis deals with the activation of such a component; the mast cell. More specifically, activation of mast cells by exogenous and endogenous danger signals is investigated. 1

14 1.1 MAST CELLS The mast cell was discovered in 1878 by the German scientist Paul Ehrlich [4]. Mast cells are most well-known for their role in allergic inflammation, where they act as effector cells which release compounds causing classical allergic symptoms. For many years, mast cells were mostly discussed in the context of allergy, but it is becoming increasingly clear that mast cells have many other functions. We now know that mast cells possess a Dr Jekyll and Mr Hyde -nature, where mast cell responses sometimes are beneficial and sometimes detrimental [5]. Thus, a more complex role for mast cells is emerging, where mast cells are more and more appreciated as important sentinel cells possessing regulatory functions. Key characteristics of mast cells include their granulation, their wide tissue distribution, their long life span, and their ability to selectively release mediators upon activation. In addition, mast cells are endowed with several types of receptors, enabling them to react to a multitude of activators. In this chapter, important aspects of mast cell biology are highlighted Mast cell distribution and heterogeneity Mast cells are evolutionary conserved inflammatory cells; early mast cells are thought to have appeared approximately million years ago [6]. Murine mast cells originate from a pluripotent, hematopoietic stem cell population in the bone marrow and human mast cells derive from a corresponding CD34 + /CD117 + /CD13 + bone marrow population [7-9]. Unlike other cells of the hematopoietic lineage, mast cells exit the bone marrow as immature precursors, circulate in the bloodstream and eventually migrate into peripheral tissues where they mature [7, 9]. Mast cells are commonly found near blood vessels and nerves in vascularized tissues throughout the body, and are especially numerous at sites facing the external environment, such as the skin, the airways and in the gastrointestinal tract [7, 10, 11]. As a result of this tissue distribution, mast cells are among the first cell types to encounter invading pathogens, and are therefore often referred to as gate-keepers, or sentinels, of the immune system. This important function of mast cells will be discussed in more detail in chapter Mast cells represent a heterogeneous cell population, where the tissue microenvironment and growth factor milieu dictates the characteristics of the mature mast cell. Mast cell heterogeneity was originally described by Enerbäck in the 1960s [12, 13]. In this work, mast cells in the intestinal mucosa and submucosa were called mucosal mast cells and connective-tissue mast cells, respectively, depending on their dye-binding properties. In humans, corresponding mast cell populations are referred to as MC T (tryptase containing) and MC TC (tryptase and chymase containing) [9]. However, this division probably represents an oversimplification, as it is more likely that there is a wide spectrum of mast cell variants, depending on localization and local microenvironment. These types of mast cells might also change during inflammation or infection [11]. Interestingly, it has even been suggested that mast cell phenotypes can be reversible under certain conditions [9, 14]. Stem-cell factor (SCF) is the most important growth factor for mast cells. However, other mediators such as interleukin-3 (IL-3), IL-4, IL-5, IL-6, IL-9 and transforming growth factor-1 (TGF-1) can also influence mast cell proliferation and survival [7, 11, 15-17]. The importance of SCF or its receptor c-kit, which is highly expressed by mast cells, is 2

15 demonstrated in mice deficient in c-kit or SCF, as mast cells are almost completely absent in these mice [18, 19] Mast cell activation and mediator release Mast cells are best known for their role in allergic reactions. Therefore, the most extensively studied mast cell reaction is immunoglobulin E (IgE)-dependent mast cell activation [20, 21]. This kind of activation is what underlies an allergic reaction, eventually resulting in classical allergic symptoms, such as sneezing and a runny nose. IgE-dependent mast cell activation is schematically depicted in figure 1. When susceptible individuals come into contact with for example birch pollen, their B cells produce birch-specific IgE antibodies (a process that also requires T cell help). These antibodies will then attach to the surface of mast cells by binding to high affinity IgEreceptors (FcRI). Upon a second exposure to birch pollen, IgE-primed mast cells are activated through FcRI-crosslinking, causing the mast cells to degranulate and thereby release mediators, such as histamine, that cause allergic symptoms. Y B T cell help Ig class switch IgE Y Y Y Y Y Y Y M Y Y Y M Figure 1. IgE-dependent mast cell activation. In this example, birch allergen (black star) induces the production of IgE antibodies from B cells (1). These birch-specific IgE-antibodies attach to mast cells by interacting with the FcεRI receptor (2). When the same birch allergen is encountered a second time, IgE-primed mast cells become activated and degranulate (3), thereby releasing mediators that elicit allergic inflammation. Aside from IgE-dependent mast cell activation, mast cells can become activated through many other means. Mast cell activators derive from a wide variety of sources, including viruses, parasites, cytokines, chemokines, chemicals, venoms, endogenous peptides, complement [21], neuropeptides such as substance P [22], IgG [23, 24] fungal components such as zymosan [25], and CD30/CD30L interactions [26]. Studying alternative or IgE-independent routes of mast cell activation is of great importance in order to better understand the different roles mast cells have in innate immune responses. While considering that there are many different types of signals capable of mast cell activation, it is important to remember that mast cells do not aimlessly release all their content when activated. Instead, mast cells release different mediator profiles depending on the triggering factor [27]. This means that a certain stimuli might cause degranulation, while another might induce release of cytokines or chemokines without triggering degranulation. For example, many pathogen products induce release of 3

16 cytokines and chemokines from mast cells, but do not induce degranulation [11]. The ability of mast cells to selectively produce and release mediators means that they can tailor responses as required for a certain situation. This is of particular importance for their role as sentinel cells, as it gives them the opportunity to fine tune mediator release depending on the invading pathogen. In the same way as mast cells can be activated by many different triggers, they also produce and release many different mediators. To simplify, mast cell mediators can be divided into two groups; i) mediators that are preformed and stored in granules and ii) mediators that are synthesized upon activation [21]. Mediators belonging to the first group are released immediately during degranulation, while mediators of the second group can be released minutes to hours following mast cell activation. Mast cell mediators are summarized in figure 2. Activator M Pre-stored Proteases Histamine Heparin Serotonin Cytokines VEGF FGF2 Lipid mediators LTB 4 LTC 4 PGD 2 PGE 2 PAF Figure 2. Examples of mast cell mediators. Cytokines & Chemokines TNF IL-1 IL-6 IL-10 IL-13 IFN- IFN- IFN- GM-CSF IL-8/CXCL8 MCP-1/CCL2 MIP-1/CCL3 MIP-1/CCL4 RANTES/CCL5 MIP-2/CXCL2 KC/CXCL1 Of the pre-stored mediators, different proteases (including tryptases, chymases and carboxypeptidases) constitute the largest group. Other pre-stored mediators include histamine, serotonin, heparin and tumor necrosis factor (TNF) [11, 21, 27]. Mediators synthesized upon activation include lipid mediators such as leukotriene B 4 (LTB 4 ), LTC 4, prostaglandin D 2 (PGD 2 ) and PGE 2 [11] and many different cytokines and chemokines, including TNF, IL-1, IL-3, IL-4, IL-5, IL-6 IL-8/CXCL8, IL-9, IL-10, IL-13, IL-16, GM-CSF, MCP-1/CCL2, MIP-1α/CCL3, IFN-α, IFN-β and IFN-γ [11, 27]. In addition, mast cells can also produce and release antimicrobial peptides such as LL-37 [28]. The ability to selectively release different mediators depending on the activating stimuli underlies the fact that mast cells participate in, and help orchestrate, many different 4

17 responses. These can be beneficial or detrimental, depending on the circumstances. The impact of various mast cell responses are discussed in chapter Mast cell function in health and disease In 2007, Mitch Leslie published a News article in Science entitled Mast cells show their might, in which Leslie stated: Once dismissed as allergy cells, mast cells have proven crucial for immunity. But they ve also shown a dark side [5]. Today, it is clear that mast cells can be beneficial, detrimental, or even fatal, depending on the current setting that they are present in, a fact that renders them particularly interesting targets of research. An overview of mast cell roles in health and disease is depicted in figure 3. In order to understand mast cell function in health and disease, the use of mast celldeficient mice has been an invaluable tool. The two most commonly used strains are referred to as WBB6F1-Kit W/W-v [29] and C57BL/6-Kit W-sh/W-sh [30]. These mice almost completely lack mast cells due to mutations in the c-kit gene, which encodes the receptor for SCF. Comparing results in wild-type mice, mast cell-deficient mice and mast cell knock-in mice has provided important knowledge concerning the role of mast cells in different diseases [31, 32]. In addition, reconstituting mast cell-deficient mice with mast cells deficient in a certain mediator facilitates studies on the role of mast cellderived mediators, for instance TNF [9, 33]. In addition to Kit-mutant mice, two novel mouse models were recently generated in which mast cells are selectively targeted and eradicated using genetic engineering [34, 35], allowing for more precise future studies of mast cell roles in the immune system. When considering the mast cell s characteristics, it is easy to understand why they are now recognized as important sentinel cells. By being pre-positioned in tissues, longlived and able to selectively release inflammatory mediators, mast cells are well equipped for being first hand-responders to different pathogens [36]. The importance of mast cells in bacterial defense has been extensively studied. For instance, mast cells have been shown to be of vital importance in acute bacterial peritonitis models [37], as well as in bacterial clearance after intraperitoneal challenge with enterobacteria [38]. In addition, mast cells protect against other bacteria, such as Helicobacter [39], Pseudomonas [40] and Mycoplasma [41]. Mast cells also seem to have important roles in the defence against viruses and fungi. For instance, mast cells are activated by poly(i:c) (a dsrna analogue), reovirus [42] and dengue virus [43]. In addition, fungal products such as zymosan and fungal extract (from Malassezia sympodialis) also activate mast cells [25, 44, 45]. Supporting a role for mast cells in the recognition of virus and fungi, mast cells express receptors important for the recognition of such pathogens, including toll-like receptor-(tlr)3, TLR7, TLR9 [46], Dectin-1 [25, 44, 45] and Mincle [44]. Pattern recognition receptors (PRRs) such as these are discussed in more detail in chapter 1.2. Even though mast cells have been shown to be able to phagocytose bacteria [38, 47], release antimicrobial peptides [28, 48], form extracellular traps [49] and even act as antigen-presenting cells [11, 50-52], one of their most important functions in host defense might be to promote inflammation. This is achieved for instance by facilitating the recruitment of effector cells such as neutrophils by releasing pro-inflammatory 5

18 cytokines, chemokines and lipid mediators. Examples of when mast cell-derived mediators are important for inflammation are numerous. For instance, TNF derived from mast cells is important for neutrophil infiltration in peritonitis [53], and mast cellderived lipid mediators such as LTB 4 and LTC 4 contribute to early neutrophil influx during bacterial infection [54]. In addition, chemokines secreted by activated mast cells, such as KC/CXCL1, play important roles in mediating neutrophil recruitment during skin inflammation [55]. Due to the same characteristics that renders mast cells efficient in recognizing and responding to pathogens, mast cells are also ideal contributors to wound healing. Indeed, mast cells have been suggested to participate in various stages of wound healing [56]. Aside from their already mentioned ability to initiate inflammation, mast cells are involved in other important aspects of wound healing as well, such as angiogenesis [57] and matrix remodeling [58]. Furthermore, early wound closure is delayed in mast cell deficient-mice, while wound closure, extravasation and neutrophil recruitment are restored upon mast cell reconstitution in these mice [59]. Mast cells are not only important in facilitating processes required for tissue repair; they are also important in actually sensing cell injury. The function of mast cells as sensors of cell injury was suggested already in 1958 by G.B. West [60], but has surprisingly not yet been completely explored. The role of mast cells as sensors of cell injury is discussed in Papers II and IV of this thesis. Beneficial Harmful Effector cell recruitment Allergy & Asthma Parasite & fungal defense Viral defense Bacterial defense Wound healing Venom resistance Homeostasis Inflammation M Anaphylaxis Autoimmunity Skin inflammation Cancer Cardiovascular disease Mastocytosis Inflammation Figure 3. Mast cells can be helpful or detrimental, depending on localization, environment, triggering stimuli, infections and inflammation and other factors. Listed here are some processes where mast cells are involved. Even though mast cell responses are beneficial in many instances, there is also a darker side to mast cell biology, most commonly linked to their involvement in allergies, anaphylactic reactions and asthma, but mast cells are also believed to be involved in diseases such as multiple sclerosis and rheumatoid arthritis (RA) [61]. In addition, mast cells are involved in the development of heart disease and cancer by, for instance, promoting atherosclerosis [62-64] and tumor growth [61], respectively. Evidence has also been presented that connects mast cells to the development of diabetes and dietinduced obesity [65]. 6

19 After allergen-specific IgE-binding and subsequent cross-linking of FcRI-receptors, mast cells are activated and release their content. This causes an immediate or early phase reaction, which has an impact on epithelial, endothelial and smooth muscle cells due to the actions of released histamine, proteoglycans, lipid mediators, tryptase and other mediators. The release of such mediators cause increased vascular permeability, mucus production and smooth muscle contraction [66]. Sometimes, this leads to an anaphylactic shock, which can have fatal outcomes. The immediate phase is followed by late-phase reactions, which are, partly, caused by mediators de novo synthesized and subsequently released by mast cells, such as leukotrienes, prostaglandins and cytokines such as IL-5 and IL-13 [9, 61, 66]. Mast cells also play a fundamental role in asthma, which is a chronic inflammatory disease of the airways characterized by bronchoconstriction, mucus secretion, edema and tissue remodeling. Mast cells contribute to these symptoms by releasing mediators that initiate and sustain inflammation in the airways [67]. Conclusively, mast cells are heterogeneous cells shaped and adapted by their local milieu. They can be activated by a multitude of triggers, and produce selective responses tailored to a certain stimuli. Mast cell activation can sometimes have detrimental effects, but they also function as important pathogen sensors, and are among the first cells to encounter invading pathogens. Aside from their longevity, their ability to rapidly produce pro-inflammatory mediators and their favorable tissue distribution, mast cells are also endowed with a multitude of receptors for sensing pathogens. Such receptors are collectively referred to as PRRs, and constitute an essential part of the innate immune system. 1.2 PATTERN RECOGNITION Germline-encoded PRRs constitute an essential part of the innate immune system, as they allow for the detection of microorganisms. This is achieved by recognizing conserved microbial structures, referred to as PAMPs [2, 68]. The discovery that PAMP recognition by PRRs is essential for the induction of immune responses, made by Charles Janeway [68], represents a key finding in immunology research. There is also accumulating evidence that PRRs are involved in the recognition of endogenous molecules released by damaged cells, so called DAMPs. Up to date, four classes of PRRs have been identified [69]. These include the TLRs, C-type lectin receptors (CLRs), NOD-like receptors (NLRs) and Retinoic acid-inducible gene (RIG)-I-like receptors (RLRs). As an undisputable testament to the importance of PRRs, the discovery of TLRs by Bruce A. Beutler and Jules A. Hoffman was awarded the Nobel Prize in Physiology or Medicine in 2011 (The prize was shared with Ralph M. Steinman for his discovery of DCs) Toll-like receptors TLRs were the first PRRs to be discovered and are thus most extensively studied. So far, 10 TLRs have been identified in humans and 12 in mice, each recognizing PAMPs derived from a wide variety of sources, including bacteria, virus, fungi and parasites [70]. TLRs are expressed both extracellularly (TLR1, TLR2, TLR4, TLR5 and TLR6) and intracellularly in endosomes and lysosomes (TLR3, TLR7, TLR8 and TLR9) [69, 7

20 71]. The TLRs that are expressed extracellularly are involved in the recognition of bacterial membrane components such as lipopolysaccharide (LPS) (TLR4), peptidoglycan (PGN) (TLR2) and flagellin (TLR5). TLR2 also forms heterodimers with TLR1 or TLR6, where TLR2/TLR1 and TLR2/TLR6 recognize triacyl and diacyl lipoproteins, respectively [69]. Intracellular TLRs are implicated in the recognition of nucleic acids such as dsrna (TLR3), ssrna (TLR7 and TLR8) and DNA (TLR9). This is of particular importance for the initiation of antiviral responses, as the triggering of intracellular TLRs often result in the production of type I interferons (IFNs). TLRs are composed of N-terminal leucine-rich repeats (LRRs), followed by a transmembrane region and a cytoplasmic Toll/IL-1R homology (TIR) domain [69]. Upon recognition of a certain PAMP, TLRs initiate signaling events starting with the recruitment of a TIR-domain-containing adaptor molecule such as MyD88, TRIF, TRAM or TIRAP. All TLRs signal through MyD88, except TLR3 which utilizes TRIF, to eventually activate transcription factors such as nuclear factor kappa-light-chainenhancer of activated B cells (NFB) and interferon regulatory factor-(irf)3 to induce production of pro-inflammatory cytokines and IFNs, respectively [70, 72]. TLRs have been mostly studied in APCs such as dendritic cells, but it is becoming increasingly clear that TLRs play important roles in a wide range of cell types, including B cells and mast cells. For instance, it has been demonstrated that mast celldeficient mice reconstituted with TLR-mutated bone marrow-derived mast cells (BMMCs) display higher mortality than mice reconstituted with wild-type BMMCs after cecal-ligation and puncture-induced peritonitis [73]. Furthermore, the TLR4 ligand LPS has been shown to induce production of several cytokines in mast cells, such as IL-1, TNF-, IL-6, IL-10 and IL-13 [73-75], while treatment of mast cells with the TLR2 ligand PGN induces TNF, IL-4, IL-5 and IL-13, but not IL-1 [75]. These types of cytokine responses have also been shown to be synergistically enhanced during IgE cross-linking [74]. It has also been shown in one study that TLR2- but not TLR4-dependent mast cell activation results in degranulation [75]. This further exemplifies that mast cell responses can be fine-tuned depending on stimuli type [75]. While many studies on TLRs and mast cells have been conducted in a murine system, it has also been described that human cord blood-derived mast cells (CBMCs) express mrna for TLR1, TLR2 and TLR6, but not TLR4 [76]. CBMCs do not respond to LPS [76] unless primed by IL-4 [77], and this also requires the presence of serum components such as CD Nod-like receptors The NLRs are cytosolic sensors of PAMPs and DAMPs, and they can be divided into three subgroups: NODs (nucleotide-binding oligomerization domain), NLRPs/NALPs (NACHT, LRR and PYD-containing domain) and IPAFs (ICE-protease-activating factor) [78]. Some NLRs form large cytoplasmatic complexes termed inflammasomes, and are involved in proteolytic activation of the inflammatory cytokines IL-1 and IL-18. Proteins of the NLR family utilize LRRs (similarly to TLRs), as well as nucleotide-binding domains (NBD) which are thought to be involved in PAMP recognition [79]. Other domains common to NLR proteins include the caspase activation and recruitment domains (CARDs) and pyrin domains (PYDs). Two of the 8

21 first NLRs to be described, and thus most widely studied, are NOD1 and NOD2 [78-81]. NOD1 and NOD2 function as intracellular PRRs and recognize PGN, a component of bacterial cell walls. NOD1 recognizes muropeptides (ie-daps), found mainly in PGN of Gram-negative bacteria, while NOD2 recognizes muramyl dipeptide (MDP), found in both Gram-positive and Gram-negative bacteria [78]. Interestingly, it was recently demonstrated that ssrna activates NOD2 [82], indicating that new NOD ligands from different sources are likely to be found in the future. Both NOD1 and NOD2 recruit the kinase RIP2 upon activation, which subsequently activates NFB, and eventually results in production of pro-inflammatory cytokines [79]. As stated above, several members of the NLR family form cytosolic multiprotein complexes called inflammasomes [83, 84]. Examples of inflammasomes are NACHT, LRR and PYD-containing protein-(nlrp)1, NLRP3 and Absent in melanoma-(aim)2, which function as sensors of endogenous or exogenous PAMPs and DAMPs. In general, inflammasomes are though to be activated by a wide range of compounds, including DAMPs (uric acid crystals, adenosine triphosphate (ATP)), PAMPs (bacteria, virus, parasites) and environmental agents (silica, asbestos, alum) [78]. It is not yet completely understood how inflammasomes are activated in vivo, but inflammasome activation is thought to be achieved in two steps [79, 85], and is best studied in the NLRP3 inflammasome (which belongs to the NALP subgroup of NLRs). The first step is though to be provided by TLR and/or NLR signaling, which initiates expression of NLRP3, as well as transcription of important components such as pro-caspase-1, proil-1 and proil-18 [79]. The second step, which is not yet fully understood, involves interactions between the activator and the inflammasome, and result in inflammasome assembly which eventually drives the release of active IL-1 and IL-18. Very little is known about the role for NLRs in mast cells responses, but it has been demonstrated that mouse mast cells express inflammasome components: Casp1 and Asc are constitutively expressed in BMMCs, while Nlrp3 and Il1b are inducible by LPS treatment [86]. In humans, it has been shown that the numbers of intestinal NOD2 + mast cells are significantly increased in patients with Crohn s disease [87]. The role of NOD1 in human CBMCs is discussed in Paper I in this thesis. Given the roles of mast cells in innate immunity, it is clear that in time, much will be learned about the role of NLRs in mast cells C-type lectin receptors The term C-type lectin was originally used to separate Ca 2+ -dependent and independent carbohydrate binding lectins. CLRs initiate responses to pathogens by recognizing mannose, fucose and glucan carbohydrate structures. Thus, CLRs can recognize several types of pathogens including virus, fungi, mycobacteria and helminths [88]. For instance, mannose-structures are present in virus, fungi and bacteria, fucose structures in helminths (and some bacteria), and glucans are present in cell walls of fungi, plants and mycobacteria. Since all pathogens express different sets of PAMPs, these PAMP profiles elicit different immune responses. This is reflected also in CLR signaling, where activation can lead to direct gene expression through NFκB activation, or to modulation of for instance TLR signaling mechanisms [88-91]. Pathogen recognition 9

22 through CLRs leads to pathogen internalization, followed by degradation and subsequent antigen presentation. Therefore, CLRs have so far mainly been studied in DCs. The mannose receptor [92] and Langerin [93] recognize high-mannose and fucose, and these were among the first CLRs to be shown to detect pathogens [91]. Since then, many different CLRs have been described, including Mincle, dectin-1, dectin-2 and DC-SIGN. Dectin-1 recognizes β-glucans in fungal cell walls, while dectin-2 binds high-mannose and α-mannans, for instance in C. albicans, M. tuberculosis and S. cerevisiae. DC-SIGN is expressed by myeloid DCs and can recognize many different pathogens such as mycobacteria, C. albicans and Leishmania spp. through mannose and fucose recognition. SIGNR3 is the closest mouse homologue of human DC-SIGN. Mincle is a CLR mainly expressed by macrophages, and recognizes α-mannans which allows for recognition of for instance C. albicans and Malasezzia spp [91]. Interestingly, Mincle is a PRR capable of responding also to endogenous danger signals. For instance, it has been demonstrated that Mincle-expressing cells respond to SAP130, a small nuclear ribonucleoprotein released by dead cells [94]. In addition, DNGR-1 (also known as CLEC9A) is a CLR with no known PAMP ligand, but binds a yet uncharacterized endogenous ligand released upon necrosis [91, 95]. Currently, the knowledge about CLRs in mast cells is very limited. However, human mast cells have been shown to express Mincle [44], and the mrna levels were upregulated upon treatment with M. sympodialis extract. Likewise, human mast cells express dectin-1 [25, 44], and respond to treatment with zymosan and A. fumigatus hyphae with leukotriene production and IgE-independent degranulation [96], respectively Retinoic acid-inducible gene (RIG)-I-like receptors RLRs are cytoplasmatic PRRs implicated in antiviral immunity as important sensors of viral RNA, but represent the least characterized class of the PRRs. RLR detection of viral components leads to signaling mechanisms eventually resulting in IFN production required for controlling virus infections. To date, only three RLRs have been identified: retinoic acid-inducible gene I (RIG-I), melanoma differentiation associated factor 5 (MDA5) and laboratory of genetics and physiology 2 (LGP2) [97]. RIG-I and MDA5 are expressed in many tissues and share several structural domains, including CARD, a RNA helicase domain, and a C-terminal repressor domain. LGP2 lacks CARD, and is thought to function as a regulator of RIG-I and MDA5 [97, 98]. RIG-I detects many different viruses, including members of Paramyxoviridae, Orthomyxoviridae, Coronaviridae and Caliciviridae, and also DNA viruses such as Epstein-Barr virus, while MDA5 responds to Picornaviridae. There are also some viruses that are detected by both receptors, including West Nile virus, reovirus and dengue virus [97]. RLR signaling is to date not fully understood, but both RIG-I and MDA5 form CARD-CARD interactions with IPS-1 (an adaptor protein) upon recognition of viral PAMPs. IPS-1 forms a signalosome complex with IRF3 and IRF7, which together with NFκB mediate production of type I IFNs [97]. 10

23 As is the case with mast cells and CLRs, the functions of RLRs in mast cells are to date largely unknown. However, some studies have emerged in which this issue is addressed. For instance, mast cells have been shown to be activated by dengue virus [43, 99]. In a recent study by St John et al., it was shown that both RIG-I and MDA5 were activated in mast cells treated with dengue virus and that this led to mast cell degranulation and de novo synthesis of TNF and IFN [99]. In addition, mast celldeficient mice infected sub-cutaneously with dengue virus displayed increased viral burdens in lymph nodes compared to mast cell-sufficient mice, suggesting mast cells to be important in fighting the viral infection. PRR-mediated recognition of PAMPs represents a cornerstone in immunology, as it is a prerequisite for innate responses to pathogens and also represents an important step in the induction of adaptive immune responses. Importantly, it is becoming increasingly clear that mast cells participate in pathogen recognition by utilizing PRRs such as those described in this chapter. While individual functions of the four classes of PRRs are becoming better and better characterized, much remains to be learned concerning their cross-talk, and how these receptors modulate and regulate each other. When a pathogen is encountered, many different PAMPs are detected by different sets of PRRs. Therefore, it is now clear that cross-talk between different PRRs is essential in order to mount proper immune responses to a given pathogen. 1.3 DAMAGE RECOGNITION While pathogen recognition through PRRs initiates inflammatory responses, pathogens are not the only cause of inflammation; damage, or trauma, is another one. Blunt trauma is a good example, where inflammation is rapidly initiated without pathogens having entered the body. The question is how the immune system can respond to injured cells, thus initiating repair mechanisms and eventually restoring tissue homeostasis? Polly Matzinger, a pioneer in the field of immunology in general, and DAMPs in particular, once stated that the immune system needs to concern itself with two questions when faced with a threat; i) to respond or not, and ii) if the answer is yes, what kind of response should be initiated [100]? During the years, different models have been proposed to answer such questions [3]. One important model was the Infectious-nonself model, proposed by Janeway in 1989 [68]. This model stated that APCs, such as DCs, are not constantly active, but are resting until activated by PAMP recognition utilizing different PPRs. Thus, this model described how APCs could discriminate between infectious-nonself and noninfectious-self [3, 101]. However, as pointed out by Matzinger, this model does not answer questions concerning why transplants are rejected, why alum works as an adjuvant (since it is non-microbial), or what induces autoimmunity. In an attempt to provide a model capable of answering such questions, Matzinger, in 1994, proposed the Danger model [102]. This model suggested that the immune system is more concerned about recognizing damagecausing substances rather than only foreign substances. Importantly, it was proposed that APCs can be activated not only by PAMPs, but also by danger/alarm signals released by cells undergoing for instance mechanical injury or injury caused by pathogen exposure [3]. 11

24 In 2000, Shi et al. found that injecting dead/dying cells provided a strong adjuvant activity, and that the adjuvant activity was present in the cytosol, releasable for instance by mechanical rupture [103, 104]. This finding emphasizes the concept underlying the Danger model, and indeed, during the last decade, many different endogenous danger signals (or DAMPs) have been characterized, providing additional validity to the danger model. Different DAMPs are discussed in chapter Endogenous danger signals A paramount function of the immune system is to monitor the health status of different cells. In this context, cell death (other than apoptotic death) is heralding trouble; a sign of alarm that needs to be recognized and responded to. Responses to injured or killed cells, aiming at repair and a return to tissue homeostasis, can be initiated by the recognition of intracellular molecules exposed as a result of cell injury or death. Thus, an important characteristic of endogenous danger signals is that they should not be released by healthy cells. Of all DAMPs characterized so far, high-mobility group box 1 protein (HMGB1) is perhaps most extensively studied. HMGB1 functions as a chromatin-binding nuclear factor, but can also be secreted and initiate inflammatory responses [105]. Importantly, release of HMGB1 by necrotic cells has been demonstrated to generate an inflammatory response [106]. Here, bone marrow cells were challenged with dead wild-type or HMGB1 -/- fibroblasts. Interestingly, while treatment with dead wild-type fibroblasts triggered a TNF response, this was not observed using HMGB1 -/- fibroblasts, or apoptotic wild-type fibroblasts, clearly demonstrating that HMGB1 released from necrotic cells initiates inflammation. In parallel, another group demonstrated similar findings, where also HMGB1-blocking reduced activation induced by necrotic cells [107]. As mentioned in chapter 1.3, Shi and colleagues demonstrated that dead cells possessed adjuvant activity of unknown identity [103]. In a later study by the same group, the cytosol of irradiated 3T3 fibroblasts was fractionated and individual fractions were tested for adjuvant activity [108]. One single fraction was responsible for most of the activity, and identified as uric acid by using mass spectrometry. In line with this finding, the adjuvant activity of necrotic fibroblasts could be decreased if the cells were pre-treated with allopurinol (an inhibitor of uric acid formation) or uricase [108]. Thus, the authors of this study identified uric acid as an important endogenous danger signal released by necrotic cells. Aside from HMGB1 and uric acid, several additional compounds of different origin have been described as endogenous danger signals. For instance, necrotic cells have been shown to release heat shock proteins (HSPs) such as HSP70 [109], and SAP130 (a ribonucloprotein recognized by Mincle) [94]. Also DNA [110], RNA [111] and ATP [112] can act as DAMPs. 12

25 It is interesting to note that several endogenous danger signals are detected by receptors previously thought to react only to foreign molecules (PAMPs). For example, HMGB1 signals through TLR2, TLR4, TLR9 and RAGE (receptor for advanced glycation endproducts) [113], self RNA-complexes can be detected by TLR7 and TLR8 [111], SAP130 by Mincle [94], hyaluronan fragments by TLR2 [114] and Granulysin by TLR4 [115]. Thus, several PPRs have over-lapping functions and are capable of eliciting inflammatory responses both in response to pathogens, but also in response to endogenous compounds released by necrotic cells. IL-1α is another endogenous danger signal released upon necrosis, and has important functions in subsequent neutrophil recruitment [116] and initiation of inflammation [117]. Interestingly, another cytokine of the IL-1-family, IL-33, has been suggested to function as an alarmin [118, 119]. IL-33 biology, its role in the immune system, and its role as a danger signal is discussed in chapter 1.4. Furthermore, IL-33-mediated mast cell responses are discussed in Papers II-IV. To summarize this chapter, endogenous danger signals are compounds exposed or released by damaged or dead cells as a consequence of infection or trauma. The release of such danger signals functions as a warning signal to the immune system, and allows for damage recognition and actions aiming at containing the damage, and eventually repairing the damaged tissue. Due to their ability to instigate innate responses as well as present antigens to T-cells and thus initiate adaptive responses, recognition and responses to danger signals have largely been studied in DCs. However, a key hypothesis in this thesis is that mast cells also are important sensors of cell injury, based on mast cell characteristics such as a pre-positioning in tissues, their ability to selectively secrete mediators upon activation and their longevity. 1.4 IL-33 IL-33 is a recently discovered member of the IL-1 family of cytokines, and signals through the T1/ST2 receptor [120]. Originally, IL-33 was thought mainly to be involved in induction of T helper 2 (Th2) responses, but many additional functions for this cytokine have been discovered. IL-33 is a potent activator of several immune cells (including mast cells), and is involved in the pathogenesis of several diseases, but it does also mediate protective functions. Interestingly, IL-33 has been shown to function as an endogenous danger signal IL-33 expression and signaling In 2005, IL-33 was characterized by Schmitz et al. [120] as a member of the IL-1 family of cytokines, together with IL-1, IL-18 and IL-1RA. Two years prior to this study, Baekkevold et al. characterized a nuclear factor expressed in high endothelial venules, termed NF-HEV [121], which was later shown to be identical to IL-33 [122]. IL-33 is expressed in many different tissues and organs, especially by structural cell types such as epithelial cells [123], endothelial cells, fibroblasts, keratinocytes [118], airway smooth muscle cells [124] and astrocytes [125]. IL-33 is also expressed in osteoblasts [126, 127], adipocytes [126, 128], monocytes [129] and pancreatic stellate cells [130]. In some cell types, pro-inflammatory stimuli such as LPS can up-regulate IL-33 mrna and protein levels [129]. 13

26 The nuclear role of IL-33 is still incompletely understood. However, it has been demonstrated that IL-33 can associate with heterochromatin in the nucleus, where it exerts transcriptional repression [122]. In addition, IL-33 has been shown to interact with the p65 subunit of NFκB, thus reducing NFκB-triggered gene expression [131]. The recent generation of IL-33 -/- mice [132] will definitively be an important tool in future studies seeking to further investigate nuclear roles of IL-33. The IL-33 receptor consists of two subunits; IL-1R accessory protein (IL-1RAcP) and T1/ST2 [133]. While IL-1RAcP is utilized also by IL-1RI [134], T1/ST2 is specific for IL-33, and was for a long time an orphan receptor until Schmitz et al. identified IL-33 as its ligand [120]. T1/ST2 was originally described as a stable surface marker expressed by Th2 but not Th1 cells, and readily used to distinguish between these cell types [135]. Besides Th2 cells, mast cells were early on shown to strongly express T1/ST2 [136]. Complex formation between T1/ST2 and IL-1RAcP is essential for functional IL-33 signalling [133]. For instance, IL-1RAcP -/- mast cells secrete reduced cytokine levels compared to wild-type mast cells after IL-33 treatment [137], and the use of neutralizing antibodies targeting IL-1RAcP abrogates mast cell responses to IL-33 [138]. The IL-33/ST2 signalling pathway involves recruitment of MyD88, IRAK, IRAK4 and TRAF6, eventually resulting in NFκB activation [120]. TRAF6 has been shown to be a crucial component in this signalling pathway, as activation of NFκB, p38 and JNK was absent when TRAF6 -/- fibroblasts were stimulated with IL-33 [139]. Additionally, JAK2 -/- fibroblasts fail to induce IκBα degradation and NFκB activation upon IL-33 stimulation, suggesting this tyrosine kinase to be of importance in the signalling mechanism of IL-33 [140]. While IL-1β and IL-18, two other members of the IL-1 family of cytokines, both require proteolytical processing by caspase-1 in order to be rendered biologically active [134], this does not seem to be the case for IL-33. Instead, full-length IL-33 (IL , ~30kDa) has been reported by several studies to be biologically active [ ], thus clearly distinguishing IL-33 from IL-1β and IL-18. This characteristic provides one explanation for the proposed role of IL-33 as an endogenous danger signal, a role discussed further in chapter Initially, caspase-1-processing was also thought to be required for IL-33 [120], until several studies demonstrated that IL-33 was a poor substrate for caspase-1 [141, 143, 144]. Consequently, proteolytical processing of IL-33 has been a much debated subject. For instance, one study reported calpain to be mediating IL-33 processing in vivo [145], while another study observed IL-33 release in macrophages treated with inhibitors of both calpain and caspase-8, as well as in caspase-1 -/- cells [146]. Furthermore, IL-33 has even been shown to be inactivated by caspase-1 [142] as well as by the apoptotic caspases caspase-3 and caspase-7 [143]. Many published studies have utilized an artificially truncated form of IL-33 (IL , ~18kDa). While this shorter form of IL-33 can activate T1/ST2 similarly to full-length IL-33 [143], it remains unclear whether the shorter form is produced naturally. 14